Dehydropolymerisation of Methylamine Borane and an N‐Substituted Primary Amine Borane Using a PNP Fe Catalyst

Abstract Dehydropolymerisation of methylamine borane (H3B⋅NMeH2) using the well‐known iron amido complex [(PNP)Fe(H)(CO)] (PNP=N(CH2CH2PiPr2)2) (1) gives poly(aminoborane)s by a chain‐growth mechanism. In toluene, rapid dehydrogenation of H3B⋅NMeH2 following first‐order behaviour as a limiting case of a more general underlying Michaelis–Menten kinetics is observed, forming aminoborane H2B=NMeH, which selectively couples to give high‐molecular‐weight poly(aminoborane)s (H2BNMeH)n and only traces of borazine (HBNMe)3 by depolymerisation after full conversion. Based on a series of comparative experiments using structurally related Fe catalysts and dimethylamine borane (H3B⋅NMe2H) polymer formation is proposed to occur by nucleophilic chain growth as reported earlier computationally and experimentally. A silyl functionalised primary borane H3B⋅N(CH2SiMe3)H2 was studied in homo‐ and co‐dehydropolymerisation reactions to give the first examples for Si containing poly(aminoborane)s.

NMR spectroscopy: NMR spectra were recorded on Bruker AV300 and AV400 spectrometers. 1 H, 13 C and 29 Si chemical shifts are given in ppm and were referenced to the tetramethylsilane and the solvent signal: THF-d8: δ  H: 1.73, 3.58, δ  C: 25 1 H} NMR spectra H3PO4 was used as external standard. The baselines of 11 B and 11 B{ 1 H} NMR spectra have been corrected and the spectra were referenced to BF3•Et2O as external standard. MS analysis: Mass spectra were recorded on a MAT 95XP Thermo Fisher mass spectrometer in electrospray ionization mode. IR spectroscopy: IR active substances were measured on an FT-IR spectrometer Alpha from Bruker in ATR mode. The vibration bands are given in cm -1 . CHN analysis: Samples for elemental analysis were prepared in the glovebox and measured on a microanalyser TrueSpec CHNS (Leco) by combustion analysis. The results of the elemental analysis are given in percent. Gas chromatography: Using a gas-tight syringe, the gas samples were taken from the reaction chamber and injected splitless into the gas chromatograph (Agilent 7890A). The gas mixture was separated by a column of the type 60/80 Carboxen 1000 (Supelco) and registered with a thermal conductivity detector. X-ray analysis: Diffraction data for 4 were collected on a Bruker Kappa APEX II Duo diffractometer. The structure was solved by direct methods (SHELXS-97) [8] and refined by full-matrix least-squares procedures on F 2 (SHELXL-2014). [9] Diamond [10] were used for graphical representations. CCDC 1961735 contains the supplementary crystallographic data for this paper. These data are provided free of charge by The Cambridge Crystallographic Data Centre. SEC analysis: SEC analysis was performed for molecular weight determination. The SEC system consists of an isocratic pump series 1200 (Agilent Technologies, US), an autosampler series 1100 (Agilent Technologies, US), a refractive index (RI) detector Dn-2010 (λ=620 nm, Bures, DE), a multi-angle laser light scattering detector (MALLS, Wyatt Technologies, US). For the separation a PL MIXED-C column (300x7.5 mm, 5 µm PSgel, Agilent Technologies, US) was used. The flow rate was 1.0 mL/min at 25°C. The calculation of the molecular weights from the MALLS detector was performed by taking into account the dn/dc of the samples, which were determined by direct injection into the RI detector of samples with varied concentration. [11] All samples (2 or 4 mg mL -1 ) were prepared in air, then filtered through Ministart SRP 15 syringe filter (PTFE-membrane, pore size: 0.2 μm) and measured immediately. The eluent consists of THF and 0.1 wt% n-Bu4NBr (Sigma). Analysis of volumetric data: Volumetric studies were done using an automatic gas buret that operates under isobaric conditions. These conditions allow for the registration of the volume change of the setup which corresponds to the amount of gas evolved during the dehydropolymerisation reaction. Details of the experimental setup were published before. [12][13][14] Kinetic analysis of the volumetric data was done using the program "Canalys". [15] At the beginning of the catalytic experiments (2 to 3 minutes) significant gas evolution was detected in THF. This was much less pronounced in toluene. The detected volume can be attributed to adjustment of the equilibrium between liquid and gas phase due to vapor pressure and thermal effects. This was confirmed by blank experiments ( Figure S1). To determine the amount of H2 in catalytic experiments the overall volume was corrected by the vapor pressure of the solvent at T = 25°C. [13] The first part (t = 3 min) was not considered for kinetic analysis.
. Volumetric curves from blank experiments using only THF or toluene.
General procedure for the catalytic dehydropolymerisation of H3B•NMeH2: H3B•NMeH2 (60 mg, 1.33 mmol) and the corresponding catalyst were weighed in the glovebox and transferred to a three-necked reaction vessel. Then, H3B•NMeH2 containing dehydrogenation vessel was connected to the gas buret under an Ar atmosphere. The gas buret was initialised and H3B•NMeH2 and the catalyst were dissolved in 4 mL of THF or toluene and data acquisition was started immediately. After completion of the dehydrogenation reaction, a gas sample was taken and analysed by GC-TCD and an aliquot was analysed by 11 B/ 11 B{ 1 H} NMR spectroscopy. The reaction solution was cannula-transferred into an oven dried Schlenk flask under Ar flow. The polymer was obtained by precipitation into 50 mL of cold (-78 °C) n-hexane, allowed to precipitate for 30 minutes and subsequently filtered. The pale yellow solid was dried in vacuum overnight. Isolated yields varied from 10 -70%.     Synthesis of N-trimethylsilylmethyl-cyclotriborazane [H2BN(CH2SiMe3)H]3: The attempt to obtain the corresponding cyclic borazane from compound 4 was carried out in analogy to the synthesis of cyclic N-methyl-borazane reported by Vaultier et al. [16] The cyclic borazane was obtained by thermally induced dehydrogenation of 4 (500 mg, 4.27 mmol). The amine borane was heated to 120 °C within 30 minutes and kept at this temperature for one hour. During cooling, a white solid precipitated consisting of the desired cyclic borazane as well as unreacted 4 and N-trimethylsilylmethyl-borazine [HB=NCH2SiMe3]3. Through sublimation at 70 °C at 1•10 -3 mbar the by-products were removed and pure cyclotriborazane remained as a white solid at the bottom of the flask. The product consists of two isomers.

Reactions in THF:
The reaction was carried out in an oven dried Schlenk flask containing Fe catalyst (15.6 mg, 0.04 mmol, 2 mol%) and H3B•NMeH2 (90 mg, 2 mmol) in 6 mL of THF. The flask was connected to a bubbler. After reaction, an aliquot of 0.4 mL was taken from the solution and transferred into an NMR tube for NMR analysis.

Reactions in toluene:
The reaction was carried out in an oven dried Schlenk flask containing Fe catalyst (3.9 mg, 0.01 mmol, 0.5 mol%) and H3B•NMeH2 (90 mg, 2 mmol) in 6 mL of toluene. The flask was connected to a bubbler. After reaction, an aliquot of 0.4 mL was taken from the solution and transferred into an NMR tube for NMR analysis.

Kinetic analysis
As can be seen from the figures below, classical kinetic analysis using linearisations of the integrated first-order rate law give significant deviations from ideal linear behaviour, especially for low concentrations of the Fe catalyst ( Figure S17). Generally, such systematic deviations can be assigned to systematic errors in [S]0 (in this case [H3B•NMeH2]0), caused by wrong weighing (problems with transfer of chemicals). For this reason, we have investigated our experimental volumetric data using non-linear regression analysis. Examples are shown below. Values determined from this analysis are given in Table 1 and Table S1. Analysis using non-linear regression (first-order) fits well for dehydrogenation reactions at high catalyst concentration ( Figure S18, left, see R.M.S. and standard deviation values). When going to lower catalyst-to-substrate ratio and high substrate concentrations (e.g. 0.5 mol% Fe, [H3B•NMeH2]0 = 0.33 M) the calculated data results are less accurate ( Figure S19, left) according to first-order evaluation. Using nonlinear regression in a Michaelis-Menten approach seems to be the best model to describe the system in these cases ( Figure S19, right). It should be noted that due to poor solubility of H3B•NMeH2 in toluene we were not able to reach the saturation range (vmax). In our case the standard reaction conditions (0.5 mol% Fe,[H3B•NMeH2]0 = 0.33 M) correspond to a range v≈vmax/2. Comparison of first-order and Michaelis-Menten model, toluene   Plotting of the reaction profiles in toluene and THF against t· [1] x shows a good visual overlay for the power value/reaction order x = 1 in both cases ( Figure S23). It should however be noted that slight deviations are observed for low catalyst concentrations. We believe that this indicates differences in the overall kinetic regime of the reaction which can be regarded as a limiting case of first-order/Michaelis-Menten kinetics.                 Polymer growth kinetics H3B•NMeH2 (60 mg, 1.33 mmol) and the corresponding Fe catalyst were weighed in the glovebox and transferred to a three-necked reaction vessel. Then, H3B•NMeH2 containing dehydrogenation vessel was connected to the gas buret under Ar atmosphere. The gas buret was initialised and H3B•NMeH2 and Fe catalyst were dissolved in 4 mL of THF or toluene and data acquisition was started immediately. The reaction was stopped at certain points and the solution was immediately transferred into a Schlenk flask containing 50 ml cold n-hexane (-78°C). The polymer was precipitated, filtered off and dried in vacuum overnight.
In situ dehydropolymerisation of H3B•NMeH2 using complex 3 in the presence of cyclohexene (NMR experiment)

Stoichiometric reaction of 3 with H3B•NMeH2
In a Young NMR tube H3B•NMeH2 (1.3 mg, 0.028 mmol) and 3 (11 mg, 0.026 mmol) were dissolved in 0.6 mL of THF-d8 or toluene-d8. No other Fe species then 3 was detected by 31 P{ 1 H} NMR spectroscopy.  # Impurities from the synthesis of complex 3. [5] Catalytic dehydrocoupling of H3B•NMe2H H3B•NMe2H (78 mg, 1.33 mmol) and the corresponding Fe catalyst were weighed in the glovebox and transferred to a three-necked reaction vessel. Then, the H3B•NMe2H containing dehydrogenation vessel was connected to the gas buret under Ar atmosphere. The gas buret was initialised and H3B•NMe2H and Fe catalyst were dissolved in 4 mL of THF or toluene and data acquisition was started immediately. After completion of the dehydrogenation reaction, a gas sample was taken and analysed by GC-TCD and an aliquot was analysed by 11 B NMR spectroscopy.  Catalytic dehydrocoupling of H3B•N(CH2SiMe3)H2  [2] 32.0 ppm (toluene-d8, br) diaminoborane HB(NRH)2 27.7 ppm (THF, d br) [2] 28.

Catalytic dehydropolymerisation of 4 with [(POCOP)IrH2] and [Rh(COD)Cl]2 in THF:
Amine borane 4 (156 mg, 1.33 mmol) and the corresponding catalyst were weighed in the glovebox and transferred to a three-necked reaction vessel. Then, the dehydrogenation vessel containing 4 was connected to the gas buret under Ar atmosphere. The gas buret was initialised and 4 and the catalyst were dissolved in 4 mL of THF and data acquisition was started immediately. After completion of the dehydrogenation reaction, a gas sample was taken and analysed by GC-TCD and an aliquot was analysed by 11 B NMR spectroscopy.  Catalytic dehydropolymerisation of 4 using complex 1: 4 (100 mg, 0.85 mmol) and the corresponding Fe catalyst were weighed in the glovebox and transferred to a three-necked reaction vessel. Then, the dehydrogenation vessel containing 4 was connected to the gas buret under Ar atmosphere. The gas buret was initialised and 4 and Fe catalyst were dissolved in 4 mL of THF or toluene and data acquisition was started immediately. After completion of the dehydrogenation reaction, a gas sample was taken and analysed by GC-TCD and an aliquot was analysed by 11 B NMR spectroscopy. Reactions in THF: An increase in viscosity was not observed and the addition of n-hexane did not result in any precipitation. The solvent was removed in vacuum and the crude residue was analysed by NMR spectroscopy.    Reactions in toluene: During the reaction the viscosity of the solution increased. The catalyst was separated from the polymer by addition of n-hexane which resulted in the precipitation of the polymer and was followed by filtration.       Co-dehydropolymerisation of H3BNMeH2 and 4: H3B•NMeH2 (30 mg, 0.067 mmol), 4 (78 mg, 0.067 mmol) and the corresponding Fe catalyst were weighed in the glovebox and transferred to a three-necked reaction vessel. Then, amine borane containing dehydrogenation vessel was connected to the gas buret under Ar atmosphere. The gas buret was initialised and the amine borane adducts and Fe catalyst was dissolved in 4 mL of THF or toluene and data acquisition was started immediately. After completion of the dehydrogenation reaction, a gas sample was taken and analysed by GC-TCD and an aliquot was analysed by 11 B NMR spectroscopy. The solution was concentrated in vacuum and acetonitrile was added, resulting in precipitation of plates, followed by filtration. The pale yellow solid was dried in vacuum overnight.